Paradoxical inhibition of cardiac lipid peroxidation in cancer patients treated with doxorubicin. Pharmacologic and molecular reappraisal of anthracycline cardiotoxicity

Anticancer therapy with doxorubicin (DOX) and other quinone anthracyclines is limited by severe cardiotoxicity, reportedly because semiquinone metabolites delocalize Fe(II) from ferritin and generate hydrogen peroxide, thereby promoting hydroxyl radical formation and lipid peroxidation. Cardioprotective interventions with antioxidants or chelators have nevertheless produced conflicting results. To investigate the role and mechanism(s) of cardiac lipid peroxidation in a clinical setting, we measured lipid conjugated dienes (CD) and hydroperoxides in blood plasma samples from the coronary sinus and femoral artery of nine cancer patients undergoing intravenous treatments with DOX. Before treatment, CD were unexpectedly higher in coronary sinus than in femoral artery (342 +/- 131 vs 112 +/- 44 nmol/ml, mean +/- SD; P < 0.01), showing that cardiac tissues were spontaneously involved in lipid peroxidation. This was not observed in ten patients undergoing cardiac catheterization for the diagnosis of arrhythmias or valvular dysfunctions, indicating that myocardial lipid peroxidation was specifically increased by the presence of cancer. The infusion of a standard dose of 60 mg DOX/m(2) rapidly ( approximately 5 min) abolished the difference in CD levels between coronary sinus and femoral artery (134 +/- 95 vs 112 +/- 37 nmol/ml); moreover, dose fractionation studies showed that cardiac release of CD and hydroperoxides decreased by approximately 80% in response to the infusion of as little as 13 mg DOX/m(2). Thus, DOX appeared to inhibit cardiac lipid peroxidation in a rather potent manner. Corollary in vitro experiments were performed using myocardial biopsies from patients undergoing aortocoronary bypass grafting. These experiments suggested that the spontaneous exacerbation of lipid peroxidation probably involved preexisting Fe(II) complexes, which could not be sequestered adequately by cardiac isoferritins and became redox inactive when hydrogen peroxide was included to simulate DOX metabolism and hydroxyl radical formation. Collectively, these in vitro and in vivo studies provide novel evidence for a possible inhibition of cardiac lipid peroxidation in DOX-treated patients. Other processes might therefore contribute to the cardiotoxicity of DOX.